Overview

This section describes graph-based quantization, including the quantization types, quantization processes, and API call examples.

Based on whether retraining is required, quantization is classified into PTQ and QAT. For details about quantization concepts, see Quantization. For details about quantization algorithms, see Compression Algorithms.

• PTQ

  Quantization takes two forms: Manual Quantization and Accuracy-based Automatic Quantization based on whether the quantization configuration file is manually tuned after quantization.

  If the accuracy of the quantized model is not as expected, perform accuracy-based automatic quantization (recommended) or Manual Tuning until the accuracy meets your requirement.

• QAT

  Currently, QAT supports quantization only for float32 network models.

  Currently, only manual quantization is supported. If the accuracy of the quantized model is not as expected, perform Manual Tuning.